AbstractsComputer Science

Infrastructure serves as the backbone of the city and hence plays a significant role in its urban structure. Therefore, it is of utmost importance to monitor its performance and assure its compliance with the growth in demand. Due to their hidden and passive nature, sewer pipelines are neglected making it essential to assess their conditions and address their associated problems to maintain quality productivity and avoid high social costs. Currently, 30% of the Canadian Infrastructure has been evaluated to be in fair to very poor conditions with a cost of 39 billion for infrastructure repair (Felio et al. 2012).In 2008, it was stated that the capital investment needs in the United States are 15 billion annually for the coming 20 years totaling to 298 billion. Moreover, the pipelines in the U.S represent 3/4th of the total needs marking the largest capital need (ASCE 2013). The current condition assessment protocols are limited to several issues including poor accuracy caused by uncertain human judgments and imprecise assessments due to the consideration of the peak score (worst defect) as the total condition score. Therefore, the development of a sound condition assessment protocol with a unified classification of distress indicators regardless of the inspector’s expertise is needed to ensure safety and quality service to the public. The objective of this research is to develop a defect-based condition assessment model as well as a protocol for sewer pipelines. This model aims to cover the structural, operational, and installation / rehabilitation defects that are associated with the pipelines, joints, and manholes of each pipe length / segment. This Fuzzy Synthetic Evaluation model consists of the Analytic Network Process (ANP) model which covers the interdependencies between the components and their defects in order to deduce their relative importance weights. The second model utilizes the defects’ severities to develop fuzzy membership functions based on a predefined linguistic condition grading scale that would precisely indicate the degree of distress. This model quantifies the distress indicators and encodes their condition linguistically (states) and numerically (scores). Furthermore, a robust aggregation model based on the Hierarchical Evidential Reasoning (HER) and Dempster-Shafer (D-S) theory is created to integrate the defects’ conditions and to evaluate the overall condition of the sewer pipeline. Also, the main grading scale in this model was developed using the K-Means clustering technique. The final condition grade is represented as a crisp value calculated by the weighted average defuzzification method. The data utilized in this research was obtained from sewer condition classification manuals, previous research, and questionnaires distributed to professionals in Qatar and Canada. Also, a sewer protocol was developed, calibrated, and verified by experts’ feedback. The fruit of this fusion was also presented in a user-friendly automated tool (SPCAT). The developed model was implemented in 29 case…